The Competitive Relationship between Japanese Accent and...

49

The Competitive Relationship between Japanese Accent and Intonation (Toshiyuki Sadanobu)Japanese Speech Communication 1　2013.3.

論文

The Competitive Relationship between Japanese Accent and Intonation

Toshiyuki Sadanobu

要旨：本稿は、現実のコミュニケーションを考慮に入れることによって、現代日本語（共通語）の話しことばに対する我々の理解を、特に語彙的なアクセントとフレーザルイントネーションの関係に関して深めようとするものである。伝統的には、共通語における語彙的アクセントとフレーザルなイントネーションは、基本的に非競合的と考えられてきた。だが、この考えでは、連なり型や重なり型は説明できるが、排他型は説明できない。排他型はこれまでほとんど取り上げられていなかったが、よく注目してみると、いくつかの方言と同様、これまで伝統的に考えられていたよりも、頻繁に幅広く観察できる。そこで本稿では、語彙アクセントとイントネーションの関係について、競合説を提案する。結論は以下 4点である。第 1点：日本語（共通語）の語彙的アクセントとフレーザルなイントネーションの関係をとらえるには、伝統的な非競合説よりも競合説の方が有効である。アクセントとイントネーションは潜在的には、ピッチへの反映をめぐって潜在的に競合している。つまり基本は排他型である。第 2点：非競合説よりも競合説の方がよいという本稿の主張は、以下 2つの観察結果により支持される。観察 1：これまで考えられていたよりもずっと頻繁に幅広く生じることが判明した排他型を説明できるのは競合説だけである。観察 2：連なり型や重なり型の場合には話し手の強い気持ちがなく、このことを説明するには競合説に立ち、「強さのアイコニシティ」の考え（話し手の気持ちが強いほど、その気持ちに対応する韻律が強い力

50


を得て、排他型を生み出しやすくなるという考え）を取り入れる必要がある。第 3点：強さのアイコニシティは、少なくとも 2点の根拠を持つ。それ

は、疑似的な最小対立を用いた知覚実験と、アクセントがイントネーションを抑え込む現象の存在である。第 4点：これまで非競合説が基本と考えられていたのはおそらく、実験室的環境では被験者はふつう強い気持ちを示さないせいだろう。

キーワード：日本語、アクセント、イントネーション、競合説、強さのアイコニシティ

Abstract：This paper aims to deepen our understanding of grammar of spoken Standard

Japanese with special reference to the relationship between lexical accent and

phrasal intonation by taking actual communication into consideration.

Traditionally, lexical accent and phrasal intonation in Standard Japanese have

been thought as non-competitive. By this view, there are no competition for

realization between lexical accent and phrasal intonation. Such a view of non-

competition can explain copulative or cumulative forms, whereas it cannot

explain conflictive form. Although research on Standard Japanese have been

reluctant to admit conflictive forms, close investigation reveals that conflictive

form is much more often and widespread than has traditionally been thought

as well as in some of the other dialects. This paper suggests a new view which

can explain all of these three pitch forms. Conclusions are as follow.

First, the relationship between lexical accent and phrasal intonation in

Standard Japanese can be better understood by competitive view rather than by

traditional non-competitive view. Lexical accent and intonation are potentially

competing for generating their own pitch form (i.e. conflictive form).

Second, the superiority of competitive view over non-competitive view

is supported by two observations: (i) It is only competitive view that can

51


explain conflictive form, which is much more often and widespread than has

traditionally been thought; (ii) By adopting Iconicity of Strength (i.e. The

stronger the speakerʼs attitude is, the more likely the corresponding prosody has strong power to generate a conflictive form), competitive view rightly

expects lack of strong attitudes in copulative and cumulative forms.

Third, Iconicity of Strength can be justified at least in two ways: one is

perceptual test, and the other is the possibility of upset victory of accent over

intonation.

And fourth, non-conflictive forms have traditionally been thought as basic,

probably because it is rare for informants to show their strong attitudes and

prominent characters in the environment of laboratory.

Keywords：Japanese, accent, intonation, competitive view, iconicity of strength

1. IntroductionIn this paper I aim to deepen our understanding of grammar of spoken

Standard Japanese with special reference to the relationship between

lexical accent and phrasal intonation by taking actual communication into

consideration. Traditionally, lexical accent and phrasal intonation in Standard

Japanese have been thought of as non-competitive. This view, however, has

many more problems than has traditionally been thought. This paper suggests

a new competitive view which can explain all three pitch forms

Previous research based on the non-competitive view is outlined in Section

2, and its defects are examined in Section 3. In Section 4 I shall show that

the competitive view is more preferable than the non-competitive one, and

in Section 5 I shall raise counter-arguments against the competitive view and

refute them. Finally I conclude my arguments in Section 6.

52


2. Previous researchTraditionally, lexical accent and phrasal intonation in Standard Japanese

have been basically thought of as non-competitive. By this view, there is no

competition for realization between lexical accent and phrasal intonation. For

example, Amanuma et al. (1978) says that intonation cannot break patterns of

lexical accents in Japanese language. Another example is Abe (1998) which

says “(Japanese lexical accent is) susceptible but usually not subservient to intonation... (It) resists being perturbed by intonation.” See the introductory part of Sadanobu (2005a) for more details of the traditional view.

According to the traditional view (“non-competitive view,” henceforth), lexical accent and phrasal intonation jointly generate either one of two pitch

forms; “copulative” or “cumulative.”By “copulative,” I mean here a pitch form where a given accent is

accompanied by intonation in this order. Let us take a word for example.

The lexical accent of the word ame (rain) is a (High) – me (Low), which can

be ascertained when we pronounce this word isolated from other parts of

utterance (Sound 1, Figure 1) (Note 1). (I use praat (praat 5207, http://www.

fon.hum.uva.nl/praat/download_win.html) for extraction and analysis of sound.

The same hereinafter unless otherwise noted.)

Sound 1: Isolated pronunciation of the word ame (rain)

Figure 1: Pitch of Sound 1

53


And the pitch of the interrogative sentence ame? (Rain?) consists of two parts.

The former part reflects the lexical pattern (High-Low) of the word ame,

and the latter part reflects the phrasal pattern (rising) of the interrogative

intonation (Sound 2, Figure 2).

Sound 2: Interrogative sentence ame?


This is why we can say that the relationship between lexical accent and phrasal

intonation of ame? (Rain?) is copulative

When the directions of accent and intonation are the same, rising for

example, a cumulative form often appears. By “cumulative” I mean a pitch form which is affected by intonation (rising), and becomes more rising. For

example, the lexical accent of the word ame (candy) is a (Low) – me (High)

(Sound 3, Figure 3).

54


Sound 3: Isolated pronunciation of the word ame (candy)


And its rising pattern becomes more remarkable in the interrogative sentence

ame? (Candy?) (Sound 4, Figure 4).

Sound 4: Interrogative ame? (Candy?)


Therefore we can say that the relationship between the lexical accent and

phrasal intonation of ame? (Candy?) is cumulative.

Copulative and cumulative forms introduced above can be explained by the

non-competitive view as a matter of course. On the other hand, there remains

a pitch pattern which the non-competitive view cannot explain by nature. This

is what we call the conflictive form (Table 1).

55


Table 1: The non-competitive view can account for copulative and

cumulative forms but not conflictive forms.

copulative cumulative conflictive

Non-competitive view + + －

By a conflictive form, I mean a pitch pattern which follows either lexical

accent or phrasal intonation. In most cases of the conflictive form, accent is

affected by intonation to such a degree that it has completely lost its original

shape. For example, the lexical accent of the word nani (what) is na (High) – ni

(Low) (Sound 5, Figure 5).

Sound 5: Isolated pronunciation of the word nani (what)


However, this word can be uttered in rising pitch with a strong attitude of

asking the question. (Sound 6, Figure 6).

56


Sound 6: Strong question nani? (What?)


In this case we can say that lexical accent and phrasal intonation are conflictive.

Although the pitch pattern slightly descends at the last part in Fig. 6, it does

not change our judgment of this pitch pattern as conflictive. This is because

such a subtle fall of pitch at the last part of the utterance can be seen in natural

data as well (e.g. oishii naa (Sound 20, Figure 10) shown in Section 4.1 below)

which we feel as conflictive.

Research on Standard Japanese has been reluctant to admit conflictive

forms. For example, Amanuma et al. (1978) pointed out the following three

patterns “V-tai,” “V/A/N-desho?,” and “V-tekudasai” as conflictive, but there remain many unclear points. Firstly, the pattern V-tai can be pronounced not

only in the conflictive form like “tabetai” but also in the copulative form like “tabetai.” So what is the difference between “tabetai” and “tabetai”? Secondly, what is the difference between conflictive “taberu deshoo” and copulative “taberu deshoo”? And thirdly, when and why do we pronounce this pattern in the conflictive form, like “tabete kudasai”? And why is it more natural to pronounce this pattern in the conflictive form with the sentence-final particle

“yo,” as in “tabete kudasai yoo”? Such questions all remain unanswered. The situation is similar for other studies.

It is not right that there has been no description that admits the competitive

relationship between lexical accent and phrasal intonation (Sadanobu 2005a).

57


Kawakami (1963: 33, 37) notes “Accentual nuclear disappears because of rising tone,” and “when the speaker utters very lightly with rising intonation (second type), s/he must delete the accentual fall.” Akinaga (1966: 58–59) admits that tones of respectful reading and of recitation and the tone of child

actor/actressʼs speaking “ignores” and “destroys” accentual patterns, although it is not clear whether these tones belong to intonation. Also Moriyama (1989:

173–174) points out that intonation can delete accentual fall with limitations

of word length and accentual patterns. Isamu Abe seems to have admitted

the conflictive form for a long time, and this paper borrows the English terms

“copulative,” “cumulative,” and “conflictive” from Abe (1998: 362). Kori (1997: 172–184) states that the intonation of a focused word can weaken the accents

of following words. Though distinct from the above research, this is also

regarded as one of the descriptions that admits that the reflection of lexical

accent onto pitch can be disturbed by intonation. However, such descriptions

that admit the conflictive form more or less are not so large in number. Are

conflictive forms so rare?

3. Some observationsTurning to dialectal research, we can find conflictive forms much more

commonly. According to Fujiwara (1997), for example, Southern Aso dialect

has a conflictive form of a high flat type in which the speaker can pronounce a

sentence “Aaga shigoto wa daa mo sen” (Nobody undertakes that work) with no lexical accent and a high flat pitch. Also Kyoto dialect has a conflictive form

of a meandering type in which a sentence such as “Sonna koto shitara aka hen yanai no” (Youʼre a fool, to do such a thing!) can be pronounced with a high pitch (for sonna, shi, a, and hen) and a low pitch (for koto, tara, ka, and yanaino).

Such a difference between research of dialects on one hand and of Standard

Japanese on the other hand might be attributed to the qualitative difference

of data observed. What is commonly observed in dialectal research and often

58


ignored in (laboratory-like) Standard Japanese research are utterances made

with a strong attitude. And by focusing on these, we can find conflictive forms

also in Standard Japanese. Below I shall show three examples.

The first example of a conflictive form in Standard Japanese is a gradual

rising type (Sound 7, Figure 7) (Note 2).

Sound 7: Emi no otete tabechau zo waan! (It will eat your hand, Emi. Wham!)


This sound is taken from the so-called “Emichan data,” a mother-daughter conversation database recorded in the latter half of the 1970s (Sugito 2005).

(Vol.14, Side B, around 49: 32) (The reason I use this database in spite of its

oldness is that it is actually not old at all, at least for its prosodic aspect.) And

here is an utterance of a mother who is trying to frighten her daughter, Emi, by

joking. At the last part she even cries an onomatopoeia Waan (Wham), a sound

mimicking something (maybe a monster judging from the context, or a big

fish judging from the conversation of Vol.11, Side B, around 15: 35) trying to

eat Emiʼs hand. Apart from this cry, the pitch of this utterance gradually gets higher and higher from around 90Hz to 110Hz in about 2 seconds. We cannot

admit any reflection of lexical accent of words such as Emi (High-Low) and

tabechau (High-Low-Low-Low) there.

The second example of a conflictive form is a gradual falling type (Video 1,

Sound 8, Figure 8).

59


Video 1: KinKi Kids

Sound 8: Hidoi yo (How cruel!) from Video 1


This is taken from a sound database of everyday conversation collected by my

research group in the mid 2000s. Here two young female speakers are talking

about a famous idol group KinKi Kids. According to my auditory impression,

the utterance “Hidoi yo” (How cruel!) gradually descends (around from 380Hz to 330Hz in Figure 8) over about 1.5 seconds. Here the lexical accent

of the word hidoi (cruel, Low-High-Low) is not reflected to pitch. (Although

the falling pitch shape of “Hidoi yo” in Figure 8 is not so clear because of overlapping with the other speakerʼs voice saying “KinKi, KinKi,” I think the idea that nothing can be said unless you have a clear reflection on the pitch

figure is unrealistic.)

60


The third and last example of a conflictive form is the “descent from initial high pitch” type (Video 2, Sound 9, Figure 9).

Video 2: After telephone conversation

Sound 9: Zenzen okke (No problem) from Video 2


This data is taken from an audio-visual database of everyday conversation

collected by my research group in the late 2000s. The video was captured by a

camera with a 360 degree visual field, and its upper and lower halves cover 180

degree angles respectively. (The leftmost end of the upper part is connected

with the rightmost end of the lower part, and the rightmost end of the upper

part with the leftmost end of the lower part.) Here we can see the interior

space of a house. On the upper field a woman stands with her face away from

61


us, talking with someone by telephone. Another woman on the lower field

is sitting with her profile directed toward us. After the first woman finishes

the telephone conversation in a polite tone, the second woman asks her

“Daijobu deshita?” (Was it ok?), and the first woman answers “Zenzen okke” (No problem!) with the initial part, ze, pronounced as high as 640Hz and the

remaining part pronounced with a falling pitch. The lexical accent of the word

zenzen (at all, Low-High-High-High) is not reflected to pitch.

In all three cases, there is no reflection of lexical accent in pitch. Thus the

conflictive form is much more common and widespread than has traditionally

been thought.

4. Competitive view and “iconicity of strength”Based on the observations above, I shall suggest a competitive view for

the relationship between lexical accent and phrasal intonation in Standard

Japanese. According to this view, the relationship between lexical accent and

phrasal intonation is basically conflictive. That is to say, they are potentially

competing with each other to generate their own pitch form (i.e. conflictive

form).

It goes without saying that the competitive view can explain conflictive

forms. So how can copulative and cumulative forms be explained?

I shall show that the competitive view can explain these non-conflictive

forms as well as conflictive forms by using the “iconicity of strength” hypothesis. This hypothesis can also characterize conflictive and non-

conflictive forms respectively.

The “iconicity of strength” hypothesis (hereafter IS) expresses the following idea: The stronger the speakerʼs attitude is, the more likely the speaker is to concentrate on producing the corresponding prosody. As a result, the prosody

has enough power to generate conflictive forms.

If we adopt IS, copulative and cumulative forms are characterized as forms

62


lacking the speakerʼs strong attitude, distinct from the conflictive form, which has such strong attitudes. Copulative and cumulative forms can be understood

as “drawn ties” between accent and intonation because of this lack of strong attitudes.

Below I shall show the validity of IS in two ways. One is a perceptual test

and the other is the fact that lexical accent is not always the loser and can

defeat phrasal intonation.

4.1 Perceptual testThe first way of showing the validity of IS is by conducting a perceptual test

using quasi-minimal pairs. By quasi-minimal pair, I mean here a pair such as

the following two utterances of “nani?” (What?), one of which is copulative (Sound 10, Figure 10) and the other of which is conflictive (Sound 11, same as

Sound 6, Figure 11).

Sound 10: nani? (What?) in copulative form


63


Sound 11: nani? (What?) in conflictive form (= Sound 6)


If these two utterances differed from each other only in pitch, we could call

them a minimal pair. Actually they are just a quasi-minimal pair, since they

differ not only in pitch but also in power, duration, voice quality etc. (There is

a big difference of duration at least.) Although the ideal pair for a perceptual

test should be a minimal pair, this is virtually unobtainable as far as we focus

on human raw sound and refrain from using signal processing technology, like

this paper. This is the reason for our compromise of using quasi-minimal pairs

here.

The subjects of the perceptual test were 30 university students in the

Kansai area and they were not paid. They listened to each utterance of quasi-

minimal pairs twice and were asked to detect the speakerʼs attitude. Their answers were made by selecting candidates of attitude already given to them.

It was announced to the subjects before the test that they could select more

than one candidate if they felt like doing so. And it was also announced to

them that they could select no candidate if they didnʼt find any suitable one corresponding with the utterance.

The result of the test was as follows: In the case of copulative nani? (What?),

as many as 29 subjects selected the candidate “asking” and three subjects selected “weak surprise.” There were no subjects who selected “strong

64


surprise.” On the contrary, in the case of the conflictive form, 28 subjects selected “strong surprise,” 17 subjects “weak surprise,” and no subjects “asking.” If we accept the common sense that of these three attitudes “strong surprise” is the strongest, “asking” is the weakest, and “weak surprise” is intermediate between them, we can understand this attitudinal difference

between copulative and conflictive forms with IS in the following way: Rising

intonation connected with strong attitudes has a power strong enough to

generate its own pitch form. (Note 3)

A similar result was gained in a quasi-minimal pair with verb stem + yoo

ending. For example, “miyoo?” (Watch?) in the copulative form (Sound 12, Figure 12) is likely to be interpreted as reasking such as “Did you say ʻwatchʼ?” (by 24 subjects) and unlikely to be interpreted as inductive such as “Shall we watch together?” (no subjects). Quite contrary to this, miyoo? (Watch?) in the conflictive form (Sound 13, Figure 13) is unlikely to be interpreted as reasking

(except 1 subject) and likely to be interpreted as inductive (by 27 subjects).

Sound 12: miyoo? (Watch?) in copulative form


65


Sound 13: miyoo? (Watch?) in conflictive form


Such an attitudinal difference between copulative and conflictive forms of

miyoo? (Watch?) can be understand by IS as follows: Compared with the rising

intonation of reasking utterances, that of inductive utterances has stronger

power for generating its own pitch form because the attitude of induction is

deontic and stronger than reasking.

We can see the same trend also in the case of quasi-minimal pairs with

verb stem + masen endings, although the connection between pitch form and

attitude is somewhat overlapping. For example, “iimasen?” (Do not say?) in the copulative form (Sound 14, Figure 14) tends to be interpreted as reasking

(“Did you say ʻI donʼt sayʼ?”, 17 subjects) and confirmation (“Is it right that you donʼt say?”, 24 subjects), rather than as strong confirmation (“You say, donʼt you?”, 5 subjects) and inducement (“Letʼs say together, 5 subjects). By contrast, iimasen? (Do not say?) in the conflictive form (Sound 15, Figure 15)

tends to be interpreted as inducement (28 subjects) and strong confirmation (20

subjects) rather than as confirmation (8 subjects) and reasking (1 subject).

66


Sound 14: iimasen? (Do not say?) in copulative form


Sound 15: iimasen? (Do not say?) in conflictive form


This difference between copulative and conflictive forms of iimasen? (Do not

say?) is quite naturally understood using IS once we admit the hierarchy of

attitudinal strength among reasking, confirmation, strong confirmation, and

inducement as the attitude gets stronger.

This is true also in the case of sentences with sentence-final particles. Let

us take a sentence “omuraisu oishii yo” (Rice omelets taste good.) with the sentence-final particle yo as an example. When this sentence is pronounced

in the copulative form (Sound 16, Figure 16), it is apt to be interpreted widely

not only as an utterance of relatively weak attitudes such as reasking “Did

67


you say ʻomuraisu oishii yoʼ?” (20 subjects) and asking “Do you know rice omelets taste good?” (20 subjects) but also as an utterance of relatively strong attitudes such as inducement “Letʼs have rice omelets together.” (15 subjects) and jokingly threatening “You will regret it if you have a rice omelet.” (13 subjects). By contrast when it is pronounced in the conflictive form (Sound 17,

Figure 17), it is interpreted only as utterance of inducement (28 subjects) and

jokingly threatening (28 subjects).

Sound 16: Omuraisu oishii yo (Rice omelets taste good.) in copulative form


Sound 17: Omuraisu oishii yo (Rice omelets taste good.) in conflictive form


68


It is noteworthy here that conflictive forms like those in Sound 17 become

unnatural when we drop the sentence-final particle yo. That is to say, the

sentence-final particle yo makes the conflictive form more natural. Although

the semantics of yo are not easy to capture, we can safely say that sentence-

final particles such as yo contribute to the expression of the speakerʼs strong attitudes as a result. Thus we can understand why sentence-final particles

facilitate the occurrence of conflictive pitch by using IS.

As well as yo, the sentence-final particle zo contributes an expression of the

speakerʼs strong attitudes and therefore enables conflictive pitch form. Let us take the sentence “emi no otete tabechau zo” (It will eat your hand, Emi.) for example. If pronounced in the copulative form (Sound 18, Figure 18),

the subjects tend to interpret this sentence as an utterance of confirmation

“Do you understand that it will eat your hand, Emi?” (12 subjects) and of threatening “I hereby threaten that it will eat your hand, Emi.” (18 subjects). And if this sentence is pronounced in the conflictive form (Sound 19, Figure

19), it is interpreted only as an utterance of threatening (27 subjects) and no

subjects interpreted this sentence as an utterance of confirmation.

Sound 18: emi no otete tabechau zo (It will eat your hand, Emi.) in copulative form


69


Sound 19: emi no otete tabechau zo (It will eat your hand, Emi.) in conflictive form


The “Emichan data” has not only zo sentences but also na sentences uttered in a conflictive pitch form. Let us see a scene where Emi and her mother are

reading a picture book together. There Emiʼs mother pretends to eat a rice ball in the book and exclaims the sentence “oishii na” (How delicious!) in the conflictive form (Vol. 13, Side B, 30: 20, Sound 20, Figure 20).

Sound 20: oishii na (How delicious!, natural utterance)


After preparing a quasi-minimal pair of the copulative oishii na (Sound

21, Figure 21) and corresponding conflictive oishii na (Sound 22, Figure

22), a perceptual test was conducted. The result was that oishii na in the

70


copulative form tends to be interpreted as an utterance of confirmation (10

subjects) and that of reasking (5 subjects) rather than that of exclamation (0

subjects), whereas oishii na in the conflictive form is likely to be interpreted

as exclamatory utterance rather as confirmation (3 subjects) and reasking (0

subject).

Sound 21: oishii na (How delicious!) in copulative form


Sound 22: oishii na (How delicious!) in conflictive form


4.2 Lexical accent can defeat phrasal intonation

IS has other evidence for its validity: The competition between lexical

71


accent and phrasal intonation for pitch reflection is not always one-sided. It is

rare but possible for lexical accent to defeat phrasal intonation. And when and

why it occurs can be understood only by using IS.

Let us take for example a scene from a TV broadcast of a motor race (F1

Grand Prix in Monaco, 2003 May 31, Kansai TV). In this scene a racer named

Montoya enters his car into his teamʼs (Williams BMW) pit to change tires and refuel (Video 3).

Please contact the author for this video.

Video 3: TV Broadcasting of Montoya’s pit stop.

At the last part of this scene, the broadcaster (Mr. Tsuneo Shiobara) makes

the utterance “Saa, hachibyou ni toyuu seishi jikan no ato, supiido seigen no moukerareta kono semai pitto roodo o ikimasu Pan Paburo Montooya no

mashiin!” (Now, after an 8.2 second stop, Juan Pablo Montoyaʼs car goes along this narrow, speed-limited pit road!) with high flat pitch pattern for

the most part (Sound 23, Figure 23. For the analysis of Sound 23, which has

much engine noise, I owe Zhu Chunyue who made Figure 11 by using SUGI

Speech Analyzer, software developed by Animo and supervised by Miyoko

Sugito. http://www.animo.co.jp/analyze/sugi/). (The racer referred to by the

broadcaster as Pan Paburo Montooya should be rightly pronounced as Hoan

Paburo Montooya (Juan Pablo Montoya), but here I transcribed his name

straight from my auditory impression of the broadcasting.)

72


Sound 23: Utterance saa, hachibyou ni toiu seishi jikan no ato,

supiido seigen no moukerareta kono semai pitto roodo o ikimasu Pan

Paburo Montooya no mashiin! (Now, after an 8.2 second stop, Juan

Pablo Montoya’s car goes along this narrow, speed-limited pit road!)


Here we can see the conflictive pitch form. A high flat intonation connected

with a strong “crying” excited attitude expels lexical accents and governs almost all parts of the utterance. However, for the part semai (narrow), there

is a clear up and down of pitch, reflecting its lexical accent (Low-High-Low)

(Figure 23, red circle). This means that lexical accent of semai withstands

and suppresses the high flat intonation. Why is such a reversal upset brought

about? By admitting IS, we can understand this in the following way: In the

preceding context the broadcaster conversed with the guest (a former F1

racer, Ukyo Katayama) and the reporter (Masahiko Kondo) about how narrow

the pit road is, and so the word semai was connected with his strong (emphatic)

attitude, which gave strong enough power to the lexical accent to suppress the

high flat intonation.

In Video 3 above, we can ascertain how often they talked about the

narrowness of the pit road before the time point of uttering Sound 23. And at

the end of Video 3, (immediately after Sound 23), we can hear that the reporter

73


Masahiko Kondo cries “deguchi mo semai!” (The exit is also narrow!)in the conflictive form, which implies that the narrowness of the pit road remained as

a potential topic while they broadcasted Montoyaʼs pit-stop.

What I discussed regarding the non-competitive view and competitive

view above is summarized in the following two points: Firstly, unlike the non-

competitive view, the competitive view can explain the conflictive form, which

is much more common and widespread (including the victory of lexical accent

over phrasal intonation) than has traditionally been thought. Secondly, by

adopting Iconicity of Strength, the competitive view rightly expects the lack of

strong attitudes in copulative and cumulative forms (Table 2).

Table 2: Explanatory powers of the non-competitive view and

competitive view

copulative cumulative conflictive

Non-competitive view + + －

Competitive view + + +

5. Anticipated counter-arguments and counter-counter-arguments

To our position of the competitive view, the following counter-argument is

anticipated: Contrary to the competitive view, conflictive forms are actually

often impossible. However strong the attitudes the speaker may hold, the

phrases yookoso (Welcome), subarashii wa ne (Itʼs wonderful, isnʼt it?), and sugosoo nante (oh, you intend to spend) for example cannot help but to be

pronounced as their lexical accents direct (e.g. High-Low-Low-Low for

yookoso, Low-High-High-High-Low for subarashii, Low-High-High-Low for

sugosoo). This indicates that the basic pitch pattern of Standard Japanese is

74


non-conflictive. This is the content of the counter-argument.

However there are at least two problems in this counter-argument. The

first problem is that this counter-argument confuses actuality and possibility.

The fact that a phrase is not pronounced in the conflictive form at the present

time does not ensure the impossibility of that phraseʼs being pronounced in the conflictive form. The counter-argument above, taking non-actuality of the

conflictive form as a sufficient condition of impossibility of this form, is infected

with a logical fallacy. The second problem is that this counter-argument

does not take into consideration the variety of speaker “characters,” social-psychological types. Henceforth I will explain these two problems in detail.

5.1 Actuality and possibilityHere I will show the counter-argument to the competitive view above

contains a logical fallacy: confusing actuality and possibility. Let us take

adjective-nai such as karakunai (not hot) as an example.

Traditionally adjective-nai has been pronounced in the copulative form only

(Sound 24, Figure 24) with attitudes of reasking (Did you say itʼs not hot?) or questioning (Is it hot or not hot?). Recently younger generation Japanese

pronounce it in the conflictive form with a strong attitude of requesting

agreement (I feel it is hot and I am sure you agree with me. Do you agree with

me?) (Sound 25, Figure 25).

75


Sound 24: karakunai? (It isn’t hot?) in copulative form


Sound 25: karakunai? (It isn’t hot?) in conflictive form


The point here is that the conflictive form of adjective-nai must have been

potentially possible, although the possibility was actualized just recently. Now

we all know that it is wrong to argue “However strong an attitude the speaker may hold, phrases of adjective-nai cannot help but to be pronounced as their

lexical accents direct. This indicates that the basic pitch pattern of Standard

Japanese is non-conflictive.” The non-existence of a conflictive form at present does not necessarily mean itʼs impossible.

Then how is adjective-nai in the conflictive form activated? This can be

understood as an analogy from verbal expressions (Sadanobu 2005a, b). As is

76


described in Section 4.1, the verbal negative form (verb-masen, and verb-nai

probably) is pronounced in the conflictive form due to strong attitudes, such

as strong confirmation or inducement. Adjective-nai in the conflictive form

connected with strong attitudes cannot be understood unless we admit IS.

5.2 Speaker’s characters (social-psychological type)The second problem of the counter-argument above is that it does not

take into consideration the variety of speakers. Indeed it is not rare that a

seemingly impossible conflictive form is actualized in accordance with speaker

“characters,” i.e. social-psychological types (Sadanobu 2010–12, 2011). Before judging a conflictive form as impossible, we must pay attention to the diversity

of “yakuwarigo” (Kinsui 2003). Let us take an utterance by Reika Ayanokoji (Photo 1), a famous figure at

Universal Studios Japan as an example (Sound 26).

Photo 1: Reika Ayanokoji at USJ

[http://www.usj.co.jp/CWY/meet_chara/profile02.html, Last retrieved:

2012 Nov. 23.]

77


Sound 26: Utterance 1 yokoso okoshi kudasaimashita (Welcome) by Reika Ayanokoji [http:// www.youtube.com/watch?v=YXFPGmPUwt4, Last check: 2012 Nov. 23.


In this utterance, a seemingly impossible conflictive form is actualized. The

part of yokoso is pronounced with rising pitch and its lexical accent (High-Low-

Low-Low) is totally neglected (Figure 26).

Here is another utterance by Reika Ayanokoji, which includes subarashii wa

ne (Itʼs wonderful, isnʼt it?) and sugosoo nante (oh, you intend to spend) that look impossible to pronounce in the conflictive form at first sight, as well as

yokoso (Sound 27).

78


Sound 27: Utterance 2 by Reika Ayanokoji

http://www.youtube.com/watch?v=pgwkatc_Bhw, Last check: 2012

Nov. 23.

And again conflictive forms are actualized with lexical accents neglected in her

eccentric and excited rising pitch (Sound 28, Figure 27, and Sound 29, Figure

28). Sound 28: subarashii wa ne (Itʼs wonderful, isnʼt it?) in Sound 27 Figure 27: Pitch of Sound 28

Sound 29: sugosoo nante (oh, you intend to spend) in Sound 27


Such a way of speaking is not special to the “acted” utterances of Reika Ayanokoji. It is more widely observed in actual utterances by speakers of the

hyperactive character.

In summary, the counter-argument against the competitive view lacks

validity firstly because it confuses actuality and possibility (5.1), and secondly

because it does not take the variety of speaker characters into consideration

79


(5.2).

6. ConclusionsThe conclusions of this paper are as follows.

First, the relationship between lexical accent and phrasal intonation in

Standard Japanese can be better understood by the competitive view rather

than by the traditional non-competitive view. Lexical accent and phrasal

intonation are potentially competing to generate their own pitch form (i.e.

conflictive form).

Second, the superiority of the competitive view over non-competitive view

is supported by two observations: (i) It is only the competitive view that can

explain the conflictive form, which is much more common and widespread

than has traditionally been thought; (ii) By adopting Iconicity of Strength

(i.e. the idea that the stronger the speakerʼs attitude is, the more likely the corresponding prosody has strong enough power to generate a conflictive

form), the competitive view rightly expects a lack of strong attitudes in

copulative and cumulative forms.

Third, Iconicity of Strength can be justified at least in two ways. One is

perceptual tests using quasi-minimal pairs, and the other is the possibility of

an upset victory of accent over intonation.

And fourth, non-conflictive forms have traditionally been thought of as

basic, probably because it is rare for informants to show strong attitudes and

prominent characters in the laboratory environment.

In order to deepen our understanding of accent and intonation, we must

expand our scope beyond the laboratory wall.

Notes1: Although a sound wave graph is often presented together with the pitch

graph, it is omitted in this paper. This is because this paper includes sound

80


itself and so bulky graphs of sound waves seem dispensable.

2: The final part zo sounds like do in some audio and to some listeners. I judged

the final part as zo mainly for the following two reasons: (i) The manuscript

of this data reads zo rather than do, and (ii) do is less common than zo (at

least for Osaka female speakers).

3: It may seem possible to divide the copulative nani? and conflictive nani? as

two separate homonymic words. This treatment, however, is not adopted in

this paper for two reasons. The first reason is that the copulative nani? and

conflictive nani? share the interpretation of weak surprise and they are too

close to divide, as is indicated by the perceptual test. If we divided them as

two separate words it would be difficult to explain this semantic closeness

between them. The second reason is that besides the case of nani? it is

widely attested that a single sentence varies its pitch forms (i.e. conflictive

or non-conflictive) in accordance with the strength of attitude. Once we

admit this general tendency, the division treatment of nani? is not necessary

any more.

AcknowledgementsThis paper is a revised version of my presentation “The competitive

relationship between Japanese accent and intonation” given at the 13th international meeting of European Associations for Japanese Studies (Tallinn

University, Aug. 26, 2011). I thank Bjarke Frellesvig, the chairperson, the

audience who gave me constructive questions and comments, and my

“Kaken colleagues,” especially Yasuko Nagano Madsen, Ryoko Hayashi, Satoshi Kinsui, Nick Campbell, Zhu Chun-yue, and the late Miyoko Sugito

for extensive advice and for helping in many other ways. My thanks also go

to the editorial committee for much advice for improving this paper. This

work was partially supported by the Ministry of Education, Culture, Sports,

Science and Technology (MEXT), Grant-in-Aid for Scientific Research (A)

23242023 and (B) 23320087.

81


ReferencesAbe, Isamu. 1998 Intonation in Japanese. In Hirst, Daniel, and Albert Di

Cristo (eds.), Intonation Systems: A survey of Twenty Languages, 360–375,

Cambridge, U.K.; New York, N.Y.: Cambridge University Press.

Akinaga, Kazue. 1966 Nihongo no hatsuon: Intoneeshon nado. Kouza Nihongo

Kyouiku, 2, Institute of Language Teaching, Waseda University, 48–60.

Amanuma, Yasushi, Kazuo Ootsubo, and Osamu Mizutani. 1978 Nihongo

Onseigaku. Tokyo: Kurosio Publishers.

Fujiwara, Yoichi. 1997 Nihongo Hougenʼonno Jissou. Tokyo: Musashinoshoin.Kawakami, Shin. 1963 Bunmatsu nado no joushouchou ni tsuite. Kokugo

Kenkyu, 16, Kokugo Kenkyukai at Kokugakuin University, 25–46.

Kinsui, Satoshi. 2003 Vaacharu Nihongo: Yakuwarigo no Nazo. Tokyo: Iwanami.

Kori, Shiro. 1989 Kyouchou to intoneeshon. In Miyoko Sugito (ed.) Nihongo no

On’in, Onsei (Book I), 316–342, Tokyo: Meiji Shoin.

Moriyama, Takuro. 1989 Bun no imi to intoneeshon. In Yutaka Miyaji (ed.)

Nihongogaku Yousetsu, 172-196, Tokyo: Meiji Shoin.

Sadanobu, Toshiyuki. 2005a Diverse relationships between intonation and

accent in Japanese prosody. Japanese Linguistics, 17, The National

Institute for Japanese Language, 5–25.

Sadanobu, Toshiyuki. 2005b Sasayaku Koibito, Rikimu Repootaa: Kuchi no

Naka no Bunka. Tokyo: Iwanami.

Sadanobu, Toshiyuki. 2011 Nihongo Shakai Nozoki Kyarakuri: Kaotsuki,

Karadatsuki, Kotobatsuki. Tokyo: Sanseido.

Sadanobu, Toshiyuki. 2010–2012 An Unofficial Guide for Japanese Characters.

[http://dictionary.sanseido-publ.co.jp/wp/author/sadanobu-e/ (Sanseido

Word-Wise Web)]

Sugito, Miyoko. 2005 On the speech database of three yearsʼ prattle and chat by baby Emi and her mother. Journal of the Phonetic Society of Japan, 9 (3),

52–57.

The Competitive Relationship between Japanese Accent and...

Documents

Transcript of The Competitive Relationship between Japanese Accent and...